< Back to overview




DIY Hurricane Backup Power How-To


A solar system bought and assembled off Amazon can be low-cost backup kept in the basement the rest of the year and then brought out and put together after the hurricane passes through.

The most recent American examples are Puerto Rico, where some customers won’t have access to power 2 months after landfall. These are just some US examples - they exist all over the world.

In this article we’ll show you how to build one, where to get the parts, and whether this system makes any meaningful energy and makes financial sense.




How-To Guide

My goal is to take care of the energy needs for me personally in the aftermath of a hurricane or superstorm. 4 main components are all we need to achieve this: A solar panel to collect, a battery to store, an inverter to convert the direct current to alternating current, and a “charge controller” to balance the three other components.



The kit

Note: Ad-blockers will block the Amazon links above, so here's the plain text links:
100W Solar Panel
MC4 Cable
Charge Controller
Alligator Clips
420Wh Deep Cycle Battery
150W Micro Inverter


Add parts to cart
total



I’m using bargain-basement parts intended for RV, marine & car usage which keeps my system cheap and mobile. The main components as found on Amazon are above.

   

I ordered the system on July 2nd, and with ground shipping the PV panel arrived July 11th from Canada, and the battery, wiring, controller and inverter arrived July 15th from Amazon USA.




Assembly

Hopefully you can adapt my system to your spatial situation pretty easily:

BEFORE STORM Cut and lay bare the end of the battery & inverter wires (battery disconnected, please). The panel’s wires are already bare on one end. Do a dry run connecting battery, inverter and panel to the controller — it should look like above.

   








AFTER STORM Detach the panel wires again and place the panel on the rooftop or wherever you get the most sunlight (…garden …balcony …).

   

Attach the panel to something in case any final gusts of wind pass through: I zip-tied the back of the panel to a cable which I fastened on both ends around sturdy roof pipes…

   

…and run the power lines back to your apartment (drop down the facade and into the window in my case).

Assemble the solar controller, inverter and battery into a tighter package. Re-clamp the wire from the panel to the solar controller and close the window. It should look as above. The charge begins!

In the evening, turn on the inverter. My lamp, computer, tablet & phone are all being powered simultaneously here on the day’s solar charge. This is a good backup until the power gets back up.




One month of off-grid living in my room

In San Francisco we get 4.26 hours of usable sunshine a day (or 1156 hours a year) according to Google Sunroof. My battery holds 420Wh (12V x 35Ah), and should be filled once a day without any shading. Actual production is an average 350Wh/day on the rooftop with real-world shading and loss.

Daily Power Production:
Theoretical: 4.26 sun hours/d * 100W solar = 426Wh/day
Actual: 350Wh/day

Daily Power Needs:
54Wh Macbook Air 13-Inch (one charge a day)
8Wh iPhone battery (one charge a day)
300Wh Space Heater (150W x 2h – our SF house has no heating)
_______

Total: 82Wh energy need per day in room (up to 382Wh)

This should be easily met by the solar system. I turn the inverter on when I get home to use AC lights and charge the Macbook through the power brick, and turn it off before I go to bed to avoid energy drain. My phone’s USB (which is direct current) can charge all night straight through the solar controller itself (which has USB ports) and doesn’t require the inverter. My laptop is a DC device and could be charged straight off the 12V battery, but I found it easier to just charge it with the AC power brick through the inverter. More info on what you can run off the system in the follow-up articlehere.

   

My traditional AC lamp is a non-optimized part of the system— I could get DC lights that run off USB to avoid inverting that energy, but have not done so thus far and prefer to just use the cute little thing. When I go to bed I’ve usually used around 30% of the energy anyway— I wish I could run a water heater, heating or fridge off this system to use the excess 270Wh of the daily energy production.




Learnings and Outlook

This project started with a simple idea: What if energy generation was a consumer electronic you could order off Amazon? I’ve learned that basic electrical knowledge and a little assembly bring us very close to this ideal: Self-contained renewable lighting and charging is achieved with a simple setup off Amazon.

   

Unless you live on an RV or a boat it doesn’t make financial sense yet (see epilogue #1). But if prices come down a little more (what a difference the last year made!) or manufacturing gets a bit more resource-efficient the scales could tip and this could be a green and fiscally sensible solution (see epilogue #2). And some time in the next decade this great little DIY system that can function as a back-up system today (see epilogue #3) could become a viable consumer electronic: a cheap personal power plant for urban renters. The vision is that everyone in the world could afford a basic system like this, to guarantee their basic power needs themselves. With a most-likely increasing amount of storms in heavily populated areas this will become ever more of a necessity.




Outlook The vision is that everyone in the world could afford a basic system like this, to guarantee their basic power needs themselves. With a most-likely increasing amount of storms in heavily populated areas this will become ever more of a necessity.

What if autonomous electric generation could be added room by room to a household (like window A/C units)? We could outfit all 7 rooms of my (shared) household for around $1400 with this system today. Because the kitchen and bathrooms have way higher energy usages (fridge, stove, water heater) than the other rooms, a more sophisticated system could mesh the batteries together (wirelessly?) to create a stronger system that sends power to the rooms that need it most.




Financial Payback & Embodied Energy

How long until it saves me money? The reason this system is so simple is because it doesn’t tie in to your apartment’s behind-the-meter electrical grid. This means the system is clean, but it also doesn’t feed into electrical heavyweights like your water heater, refrigerator and washer/dryer. It does charge anything you plug into it though. So can the system save me money? Back of the envelope:

Financial payback period for 100W system
System cost : $211 on Amazon at time of writing
Yearly energy creation: 365d * 4.26hsun/d * 100W = 155’490Wh/y
Yearly value creation: 155kWh/y * 15.34c/kWh = $24/y energy created
100W system payback period: $211 / $24 = 8.5 years until payback

The financial payback of the system is 9 years including battery, which is in line with many rooftop systems but doesn’t include servicing. This could be reduced to 6.5 years by adding a second 100W solar panel:

Payback period for 200W system
200W System cost: roughly $300 on Amazon at time of writing
Yearly energy creation: 365d * 4.26hsun/d * 100W = 310’980Wh/y
Yearly value creation: 311kWh/y * 15c/kWh = $48/y energy created
200W system payback period: $300 / $48 = 6.5 years until payback

Note however that after 8 years of daily use the lead-acid deep discharge battery will be spent, which I’m not taking into account here. Either way you cut it, this is not a money saving machine. Energy prices are just too low.




How green is it?

Does it have an impact on my CO2 footprint? Back of the envelope:

Production footprint PV multicristalline:
4200kWhee/kW [1] * 0.1kW = 420kWh embodied energy


Production footprint lead-acid battery:
321kWhee/kWh [1] * 0.5kWh = 161kWh embodied energy

Total Footprint: 581kWh

Annual energy production system: 155kWh/y
Payback period: 581kWh / 155kWh/y = 4+ year footprint payback
[1] http://renew.org.au/articles/energy-flows-how-green-is-my-solar/

An eventual product would use Lithium-Ion batteries once they come down in price, which have a way better energy footprint. Lead-acid batteries are used for now because they’re cheap. So no, we’re not saving any CO2 emissions here until after 4 years — not a green machine.

Energy independence and resilience In case of brownouts or blackouts, this would be a helpful way of wirelessly charging communications devices without the grid. A 200W system could even keep a small 60W, 12V refrigerator cool enough to conserve food (14hours of operation/day, cooling down 32 degrees below ambient temp). If energy prices increase (double? triple?) due to unforeseen events in the future, the financial perspective may even make sense with payback periods decreasing to four or even 2.8 years for the 100W system or for the 200W system to 3 years or 2 years, respectively. At under 2 years payback period, we’d be in similar consumer territory to 2-year phone contracts. Let’s hope components continue to get cheaper! (or energy prices increase — but I’d rather not hope for that)




Tech Specs


Tech Specs

Polychristalline Solar Panel
Size: 39.8 x 1.2 x 26.6 inches
Weight: 19.6 pounds
Voltage: 12V DC
Max power rating: 100W
Max amperage: 8.3A


Inverter
Max power rating: 150W
(turns off if you use more than this, but a 300W version is $6 more)
Voltage: 12V DC in, 110V AC out
Only DC 12V, NOT for DC 24V
USB Output: 5V/1+2.1A (5W and 10.5W respectively)


Solar controller
Voltage: 12V/24V
Ports: 2x USB, 1x DC in, 1x DC out
Charge Current: 20A
Discharge Current: 20A
Max in: 20A * 12V = 240W
Max out: 20A * 12V = 240W
USB Output: 5V/3A (15W) each
Type: PWM
Weight: 0.25 pounds
Size: 5.9 x 3 x 1.4in
Will upgrade to a USB-C version when available


Marine lead-acid battery (Deep Cycle)
Voltage: 12V DC
Storage Capacity: 0.42Kwh, 35Ah
Size: 6.5 x 7.7 x 5.1 in
Weight: 25 pounds


Cables (aka “Tender” or “B.O.M.”)
Alligator Clips (for connecting the battery to the solar controller)
20Ft Extension Cable




< Back to overview